|
Charter - s.15 Discrimination - Test. Quebec (Attorney General) v. Kanyinda
In Quebec (Attorney General) v. Kanyinda (SCC, 2026) the Supreme Court of Canada partially allowed an appeal, this brought against a Charter s.15 Quebec CA order "that refugee claimants with a work permit and residing in Quebec be “read in” to s. 3 subpara. 3 of the RCR" ['Educational Childcare Act - Reduced Contribution Regulation'], meaning that the appellant was eligible for childcare subsidy despite her ungranted refugee status.
Here the court sets out (in 2026) the Charter s.15 ['discrimination'] test:(2) The Section 15(1) Test
[47] The test for establishing a s. 15(1) infringement is well established and protects substantive equality by targeting both direct and indirect discrimination.
[48] To prove an infringement of s. 15(1), a claimant must show the challenged law or state action: (1) on its face or in its impact, creates a distinction based on listed or analogous grounds; and (2) imposes burdens or denies a benefit in a manner that has the effect of reinforcing, perpetuating, or exacerbating disadvantage (see Sharma, at para. 28; Fraser, at para. 27; Alliance, at para. 25; Centrale, at para. 22).
[49] Importantly, in cases of adverse effects discrimination, evidence that establishes a disproportionate impact on a protected group at the first stage of the s. 15(1) test will often also be relevant at the second stage, when determining whether a challenged law or state action perpetuates a group’s disadvantage. This Court has recognized that the two steps are not “watertight compartments” or “impermeable silos” (Sharma, at para. 30; Fraser, at para. 82).
(a) Step One
[50] At the first step of the test, the claimant must show that the challenged law creates a distinction based on listed or analogous grounds, either on its face or through its impact. Where a measure does not explicitly distinguish based on a protected ground, the claimant can still show that the law creates a distinction in effect because of the law’s disproportionate impact on the protected group (Sharma, at para. 40; Fraser, at para. 52).
[51] No rigid template governs how a disproportionate impact at step one is established (Fraser, at para. 55). While the claim must be clearly anchored in a recognized ground of discrimination, the claimant may also point to additional characteristics that contribute to establishing a disproportionate effect on the claimant group. In Fraser, for example, Abella J. highlighted the significance of a claimant group’s broader situation for determining a disproportionate impact at step one: “Courts will benefit from evidence about the physical, social, cultural or other barriers which provide the ‘full context of the claimant group’s situation’” (para. 57 (emphasis added), quoting Withler, at para. 43, and citing para. 64). In other words, consideration of the intersecting identities and circumstances of the claimant group can play a role in the analysis at step one of the s. 15(1) test. The analysis must be attentive to the situation and complexity of the claimant group, acknowledging that a challenged provision may operate within a context of intersecting realities, even though the disproportionate impact must be established for a particular enumerated or analogous ground. Where discrimination affects only a subgroup, as in this case, the failure to account for intersecting identities in step one of the s. 15(1) inquiry could effectively obscure the specific adverse effects experienced by that subgroup.
[52] Two types of evidence can help a claimant establish these impacts on them and the group to which they belong: (a) evidence about the claimant group’s situation, including physical, social, cultural, or other barriers faced by the claimant group; and (b) evidence about the results produced by the challenged law in practice (Sharma, at para. 49; Fraser, at para. 56). Ideally, a claimant will lead evidence from both categories. Still, no specific form of evidence is required and the evidence’s significance will differ depending on the unique facts of each case (Fraser, at paras. 61 and 67; Sharma, at para. 49).
[53] Evidence in the first category may include a claimant’s own testimony, an expert witness, or judicial notice (Fraser, at paras. 55-57). This evidence can help illustrate the unique circumstances of the group to which the claimant belongs, including their intersecting realities and identities, and how the challenged law creates a distinction based on those circumstances. For example, in Fraser, evidence was provided on the disadvantages faced by women in the workforce due to the historical and current unequal distribution of childcare responsibilities between women and men. This helped explain why and how the challenged law disproportionately impacted women (para. 98).
[54] Evidence in the second category can take different forms, including statistics, expert testimony, case studies, or qualitative evidence (Sharma, at para. 49). Evidence of this nature can help quantify the alleged disproportionate impact and make it concrete (Fraser, at para. 58). However, because of the diverse circumstances which can ground a s. 15(1) claim, and the evidentiary hurdles and asymmetry of knowledge often possessed by claimants when compared to the state, statistical evidence is not a requirement (Sharma, at paras. 36 and 49; see also Fraser, at para. 59).
[55] Furthermore, in some cases, a court may reasonably infer an adverse effect on a claimant group. The disproportionate impact on a claimant group may be “apparent and immediate” and evidence of a specific kind may not be necessary (Taypotat, at para. 33; see also Fraser, at para. 61; Sharma, at para. 49). For example, a rule or requirement may create an obvious adverse effect for most or all members of a claimant group (see, e.g., in a human rights complaint, Simpsons-Sears). Judicial notice can also be taken about a claimant group’s situation when appropriate (Fraser, at para. 57; see also Eldridge v. British Columbia (Attorney General), 1997 CanLII 327 (SCC), [1997] 3 S.C.R. 624, at para. 69).
[56] Finally, the following are not required to satisfy the first step of the s. 15(1) test:. A claimant need not explain why the law creates a disproportionate impact (Sharma, at para. 46; Fraser, at paras. 63 and 70).
. The claimant need not show that the law is the “dominant cause of the disproportionate impact” (Sharma, at paras. 45 and 49).
. There is no need to show that the legislature intended to cause the distinction (Fraser, at para. 69; Andrews, at pp. 173-74; Eldridge, at para. 62; Vriend v. Alberta, 1998 CanLII 816 (SCC), [1998] 1 S.C.R. 493, at para. 63; Alliance, at para. 28; Centrale, at para. 35).
. It has long been unnecessary for a claimant to identify a specific comparator group (Withler, at paras. 60 and 63; Sharma, at para. 41).
. There is no need to show that all members of a claimant group will be affected by the challenged measure in the same way; there may be a disproportionate impact on only some of the claimant group (Brooks, at pp. 1247-48; Quebec v. A, at para. 354; Fraser, at para. 72). [57] Some of the interveners to this appeal contend this Court’s recent decision in Sharma has altered aspects of the s. 15(1) test by revising the causation requirement under the first step (see, e.g., I.F. (British Columbia Civil Liberties Association); I.F. (Attorney General of Ontario); I.F. (Attorney General of Alberta)). This idea has been echoed by some academic commentators (see, e.g., J. Watson Hamilton and J. Koshan, “Sharma: The Erasure of Both Group-Based Disadvantage and Individual Impact” (2024), 115 S.C.L.R. (2d) 113, at pp. 119-21; C. Salvino, “R v Sharma’s ‘Clarification’ of the Section 15 Framework and its Creation of Unique Barriers for Disability-Based Equality Claims” (2024), 32:4 Const. Forum 1). The majority in Sharma stated that it wished to clarify an uncertainty regarding “causation and its relationship with the evidentiary burden to establish disproportionate impact” (para. 35). To this end, they held that “[s]ection 15(1) claimants must demonstrate that the impugned law or state action created or contributed to the disproportionate impact on the claimant group at step one” (para. 45 (emphasis in original)).
[58] What divided the majority and the dissent in Sharma was whether the claimant had met the evidentiary burden at step one: i.e., whether the legislated removal of the availability of conditional sentences for certain serious offences had a disproportionate impact on Indigenous offenders because the removal prevented recourse to the remedial Gladue sentencing framework (R. v. Gladue, 1999 CanLII 679 (SCC), [1999] 1 S.C.R. 688). The dissent found that, given the connection between the statutory provisions removing access to conditional sentences and the requirement to consider all available sanctions other than imprisonment, with particular attention to the circumstances of Indigenous offenders, the disproportionate impact on Indigenous offenders could be reasonably inferred (Sharma, at para. 227). The majority, by contrast, did not consider such a link sufficient, on its own, to infer a disproportionate impact and establish that the law created a distinction. They concluded that in finding a s. 15(1) breach, the Court of Appeal for Ontario “failed to identify any evidence” that the law “created or contributed to a disproportionate impact on Indigenous offenders” (para. 74 (emphasis in original)).
[59] Despite these differences in application, the majority in Sharma explicitly stated that they were not altering the applicable test (paras. 33 and 46). The language of causation used in Sharma simply focused on the burden that must be met at the first step of the s. 15(1) test. Causation is inherent in this first step, which asks whether the challenged law or state action creates a distinction (Fraser, at para. 27). While the majority emphasized the importance of causation, they did not change the test. They recognized that a causal link can be drawn through “a reasonable inference” and that it “may be obvious and require no evidence” — a recognition consistent with decades of this Court’s jurisprudence holding that the burden at this stage must not be unduly onerous (Sharma, at para. 49). A s. 15(1) claimant need not show that the challenged law is the sole or dominant cause of the disproportionate impact; it is sufficient for it to be a cause (paras. 44-46; see also Fraser, at para. 71). Moreover, Sharma confirmed the principle that those alleging a breach of their s. 15(1) rights need not show that the protected characteristic “caused” the disproportionate impact (paras. 46-48; Fraser, at para. 70).
[60] In conclusion, this Court has repeatedly emphasized that the first step of the s. 15(1) test is not meant to be “an onerous hurdle designed to weed out claims on technical bases” but rather “to exclude claims that have ‘nothing to do with substantive equality’” (Alliance, at para. 26, quoting Taypotat, at para. 19). To achieve the promise of substantive equality, in line with the purposive interpretation of s. 15(1), this Court has eschewed rigid evidentiary requirements in favour of a more flexible approach, regarding how adverse effects may be shown. That said, while the evidentiary burden at the first step should not be undue, it must be fulfilled (Sharma, at para. 50). A provision must be shown to create or contribute to a distinction through its disproportionate or differential impact on a protected ground.
(b) Step Two
[61] At the second step of the test, an applicant must establish that a law imposes burdens or denies a benefit in a manner that has the effect of reinforcing, perpetuating, or exacerbating disadvantage (see Sharma, at para. 28; Fraser, at para. 27; Alliance, at para. 25; Centrale, at para. 22). Not all distinctions created by law will be found discriminatory (Andrews, at p. 182; Sharma, at para. 51). This second step focuses the s. 15(1) inquiry on the “protection of groups that have experienced exclusionary disadvantage based on group characteristics” (Fraser, at para. 77). Therefore, even if a law treats people differently based on a protected ground, it does not necessarily follow that the law increases or reinforces a disadvantage.
[62] This Court’s decision in Weatherall v. Canada (Attorney General), 1993 CanLII 112 (SCC), [1993] 2 S.C.R. 872, illustrates how a law that causes a differential impact may still fail to meet the requirements of step two. There, a male inmate argued that it was discriminatory for male prisoners to be subject to frisk searches and surveillance by female guards, while female inmates were not subject to the same treatment by male guards. This Court disagreed, emphasizing that equality does not always mean identical treatment — rather, in certain cases, different treatment may be necessary to promote substantive equality. This Court found that although this was a sex-based distinction, it was not discriminatory when understood in context. As La Forest J. observed, “[t]he reality of the relationship between the sexes is such that the historical trend of violence perpetrated by men against women is not matched by a comparable trend pursuant to which men are the victims and women the aggressors” (p. 877). Because of these historical and social realities — including the fact that women have generally occupied a disadvantaged position in society — cross-gender searches affect female inmates differently, and often more severely, than their male counterparts.
[63] At step two of the s. 15(1) test, it is also necessary to recognize that discrimination cannot be neatly packaged into a single ground and a person will often live with other circumstances, realities, or identities that may enhance or exacerbate their disadvantage. An intersectional approach may play a key role at this stage by allowing a more nuanced understanding of the nature and intensity of the disadvantage experienced. In the words of two interveners: “People experience discrimination as whole persons, not as an aggregate of separate characteristics” (I.F., National Association of Women and the Law and David Asper Centre for Constitutional Rights, at para. 5). Intersectionality enriches the analysis of discriminatory effect by highlighting these additional dimensions of disadvantage. This approach also grounds the analysis in the concrete lived realities of marginalized individuals. To appreciate the effects of the distinction established at step one on the claimant or their group, it may be necessary to consider how these experiences speak to intersecting realities that make an experience of discrimination based on the identified ground particularly acute. For example, race, sexual orientation, disability, age, socioeconomic status, poverty, marginalization, or an inability to integrate into society, may increase their disadvantage.
[64] Although this Court in Sharma found the s. 15(1) claim failed at step one, the majority reaffirmed this Court’s commitment to contextual analysis in step two of the s. 15(1) test, which includes considering not only the challenged legislation but also the broader social, political, and legal context (paras. 56-61; see also Eldridge, at para. 55, citing R. v. Turpin, 1989 CanLII 98 (SCC), [1989] 1 S.C.R. 1296, at p. 1331; Vriend, at para. 106).
[65] The broader legislative context may also be relevant to the analysis at step two (Sharma, at para. 56). The relevance and weight of the legislative context will depend on the nature of the case. For example, Wagner C.J., in R. v. C.P., 2021 SCC 19, [2021] 1 S.C.R. 679, found it necessary to put more weight on legislative context given that the provision at issue in the Youth Criminal Justice Act, S.C. 2002, c. 1, formed part of a comprehensive, “multifaceted legislative scheme” that sought to balance the competing needs of the claimant group (para. 144). However, while it may be relevant to examine how a scheme seeks to balance competing legislative objectives, benefit a protected group, or allocate limited resources, these considerations are only one aspect of a broader analysis on whether a provision has a discriminatory character. In other words, the emphasis on context must not conflate an analysis of the discriminatory impact of a measure on a disadvantaged group with a consideration of whether the distinction is justified based on legislative objectives (Fraser, at paras. 79-80). As Wagner C.J. emphasized in C.P., justifying a provision based on whether it “is relevant to a legitimate state objective or on the basis of the ameliorative effect” of the scheme at issue is “properly left to the inquiry under s. 1 of the Charter, and then only to the extent that they can specifically justify the impugned limitation” (para. 153, citing Fraser, at para. 69). To emphasize legislative context in a way that blurs this line risks improperly shifting the analysis under s. 15(1) towards an implicit balancing of government interests, an inquiry better left to s. 1. The analytical distinction between finding a Charter right infringement and justifying this infringement under s. 1 has long been part of this Court’s jurisprudence (Andrews, at p. 178; see also Oakes, at p. 134; Centrale, at para. 35; Fraser, at para. 79).
[66] Although the state can act incrementally to address inequality (Sharma, at para. 64), ameliorative legislation protected by s. 15(2) is not categorically immune from s. 15(1) scrutiny (Fraser, at para. 69; Centrale, at paras. 8 and 35; Alliance, at paras. 32-33). Courts should not ignore infringements of s. 15(1) merely because the law is ameliorative and only perpetuates the disadvantage of “some” members of the protected group. As this Court held in Eldridge, “once the state does provide a benefit, it is obliged to do so in a non-discriminatory manner” (para. 73).
[67] Finally, at the second step of the s. 15(1) test, there is no need to show that the challenged measure stereotypes or causes prejudice towards the protected group (Fraser, at para. 78). In Quebec v. A, Abella J., writing for a majority on this issue, clarified these were not additional requirements at step two because such an approach incorrectly focuses the inquiry on whether a discriminatory attitude exists as opposed to a discriminatory impact (para. 327). The key question under step two is whether the law worsens or reinforces the disadvantage experienced by the protected group.
[68] Nor is it necessary to prove that the distinction is arbitrary (Fraser, at para. 80). In other words, an infringement is still made out in circumstances where there is a rational connection between a disadvantage and legitimate state objective (para. 79). An inquiry under s. 15(1) centres on the effects, rather than the purpose, of the challenged measure on a protected group (para. 41; Andrews, at p. 174).
|
